exploiting-type-juggling-vulnerabilities

Pass

Audited by Gen Agent Trust Hub on Apr 9, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill does not contain any malicious patterns, obfuscation, or data exfiltration logic. Its content and behavior are consistent with its stated purpose of security research and vulnerability testing.
  • [COMMAND_EXECUTION]: The Python script scripts/agent.py and the documentation in SKILL.md include functionality to perform HTTP POST requests to external URLs provided by the user. These network operations are intended for vulnerability verification and are controlled by the user.
  • [METADATA_POISONING]: A minor discrepancy is noted between the author name in SKILL.md ('mahipal') and the author in the LICENSE file ('mukul975'). This appears to be a documentation oversight and does not pose a security risk.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 9, 2026, 10:28 AM